ggplot混合颜色与形状映射时，如何合并图例并修改标签（无需手动设置）

Great question! You're right that when mapping the same categorical variable to both colour and shape aesthetics, ggplot2 will create separate legends unless the scale settings (name and labels) match perfectly.

The good news is there's a direct equivalent to scale_color_hue() for shapes: scale_shape_discrete(). Just like scale_color_hue(), it uses ggplot2's default shape palette, so you don't have to manually specify shape values—you only need to set the labels parameter to match your color scale's labels.

Adjusted Code to Merge Legends #

Here's how to tweak your existing code to get a single combined legend:

ggplot(subdata, aes(x=factor(rGPUs), y=Speedup, colour=factor(Factor), shape=factor(Factor), group=Factor, ymin=0)) +
  geom_line(size=1) +
  geom_point(size=4) +
  scale_color_hue(labels = c("1X", "1.5X", "2X")) +
  scale_shape_discrete(labels = c("1X", "1.5X", "2X"))

Optional: Customize the Legend Name #

If you want to rename the combined legend, just ensure both scales use the same name parameter—this guarantees they merge seamlessly:

ggplot(subdata, aes(x=factor(rGPUs), y=Speedup, colour=factor(Factor), shape=factor(Factor), group=Factor, ymin=0)) +
  geom_line(size=1) +
  geom_point(size=4) +
  scale_color_hue(name = "Factor Multiplier", labels = c("1X", "1.5X", "2X")) +
  scale_shape_discrete(name = "Factor Multiplier", labels = c("1X", "1.5X", "2X"))

Why This Works #

The core rule here is: when two aesthetics are mapped to the same variable, and their scales have identical name and labels, ggplot2 automatically combines them into one legend. scale_shape_discrete() is exactly the "no-manual-values" shape scale you're looking for—it behaves just like scale_color_hue() but for shapes, leveraging ggplot2's built-in discrete shape palette.

If you don't specify a custom name, both scales will default to the name of your mapped variable (e.g., Factor), so matching the labels alone is usually enough to merge the legends.

内容的提问来源于stack exchange，提问作者Bub Espinja